Conventionally, Alzheimer's disease (AD) and other dementias are diagnosed using clinical assessment, neuropsychology and also structural neuroimaging, showing neuronal degeneration starting in the hippocampal regions. However, there is an increasing need for a new method that is more sensitive to early AD identification than currently possible. A new promising technique that may be used for this is to measure local brain activation using functional magnetic resonance imaging (fMRI), since functional loss predates structural loss of brain tissue. A new method to apply fMRI is to study connectivity between brain regions during a resting state without application of a task. Recent data suggest that connectivity within memory systems during such a resting state is associated with the level of memory function. Here we explain how we will study healthy elderly controls, patients with a mild cognitive impairment (MCI, considered to be a transitional stage between normal condition and AD), and AD patients using resting state connectivity fMRI. If resting state connectivity is sensitive to cognitive decline, this will be of great importance for noninvasive dementia research, offering a tool to easily study functional networks in the brain without the requirement of a memory task, and perhaps offering a tool sensitive for early diagnostics.